Jim Dearing, is an IHS market analyst specialising in security and building technologies with a heavy emphasis on intrusion, access control, residential security and the smart home.  He has explored what is holding up the widespread integration of biometrics into access control systems.

Advantages of biometrics

Biometrics offer a much more personalised and more secure system. The security of enrollment is higher with biometrics than cards as the credential is linked to a user at enrollment so the user must always be present when requesting access. This is particularly sought after in installations when end users are attempting to reduce credential sharing between those entering a building. Due to the lack of a physical card being assigned, there is far less risk of credentials being lost or duplicated. It is nearly impossible to lose a fingerprint or request a new one, duplicating a face without arousing suspicion is also significantly more difficult than cloning a proximity card. In addition, the end-user does not have to worry about the monetary cost of reproducing new credentials; this is particularly relevant to sites where potentially thousands of credentials need to be assigned.

Where is the biggest uptake?

Financial intuitions are by far the biggest users of biometric technology and seem to be the most progressive when it comes to new technology adoption. Frictionless biometric access has had the biggest interest from IT and finance users but is also popular in the healthcare sector where users wear surgical scrubs and only eye is not covered.  Growth of the market for biometrics, mostly fingerprint readers, in the manufacturing and industrial sector has also increased in recent years as these systems are introduced for time and attendance functions and access control. There are now projects launched that are completely designed using biometrics as the primary credential type.

The educational sector access control market has started to see the uptake of biometrics. Access control technological developments at the university level are aimed at showing how progressive the campus is in order to boost applications. In the biometric reader market, education systems originally started with standard fingerprint readers but are increasingly moving towards finger vein technologies. Access control in the educational sector is now the third biggest market for biometrics in industry size and growth in North America.

What are the barriers to greater adoption?

The most common biometric reader type, fingerprint, is also the most exposed to spoofing and hacking. Fingerprints can be taken from anything the credential owner has touched, then copied onto an overlay and placed on the false user’s finger. Fingerprints are not usually noticed by other users, it is difficult for them to determine which set actually belongs to the user without active comparison.

Biometric readers are also typically designed to avoid false positives at all costs. This means a biometric reader would rather deny the correct credential multiple times, rather than accept an incorrect credential just once. This is ideal in high security environments, however it can also be very inconvenient in high traffic entrances where security is less of a concern.

Fingerprint readers in particular suffer from uneven credential quality across a normal distribution of users, meaning some fingerprints are better than others. In an installation where there are 200 people with access rights the biometric reader is only valuable when it can consistently and accurately read the user with the lowest profile fingerprint. This is because if it cannot read that print, the user will be given a card for access and then why not just give everyone a card?

This leads on to the other problem that biometric readers face, door traffic flow speed. The average time that an end user with high quality fingerprint can get through entrance with a biometric reader is around 2 seconds. However, with a proximity card this time is much lower, between 0.5 and 1 second. This does not seem like a lot but it soon adds up and in high traffic doorways or buildings where there are many doors this can soon become a problem for end users.

Privacy concerns regarding sharing such personal data and the price of biometric equipment are also stifling adoption. In 2015 the average selling price of a biometric reader in the Americas was more than double that of a smart card counterpart and nearly four times as expensive as a proximity reader.

What can be done to increase adoption?

The majority of the barriers currently stifling greater adoption of biometrics in access control systems revolve around either the reader failing to read a credential efficiently enough or its price being too high. As the technology improves and manufacturers are able to lower their prices some of these concerns will be alleviated naturally. However, manufacturers want to drive up adoption now and the best way to do that to find new use cases or promote new biometric reader types that are better suited to old use cases.

Frictionless technologies are forecast to continue to gain momentum during the next five years due to their superiority convenience in most installations types and their safety benefits in hazardous installations. These technologies will be driven by greater iris-recognition use in the high-security sector in the short term, but facial recognition will see greater adoption nearer 2020 because of its ability to achieve mass throughput, superior convenience and complete lack of health concerns.

Finger-vein authentication, unlike fingerprint and other forms of biometric readers (which scan the exterior of the body), has finger-vein readers that scan information in the interior of the body, thereby making falsification extremely difficult. This eliminates the security concerns surrounding the use of fingerprint readers and can easily be retrofit in their place.

In addition, privacy concerns can only truly be eliminated though education by manufacturers and installers to reassure consumers that these new systems are secure and their own data is safe. For example, when a fingerprint template is created for a system including biometric readers it is converted into binary data. This data cannot be read by other readers and also cannot be converted back into a fingerprint. But not many end-users are unaware of this and assume once their print is taken it is at risk of being shared in the public domain.

Frictionless biometrics the way forward

Dearing summarises his findings. "Biometrics will likely remain complementary products in many projects. These readers are currently too slow, expensive and inconvenient for sole use within projects where there are large numbers of doors to secure. However, they are perfect for securing restricted areas within a building where few people have access rights. In the future, improvements to frictionless biometric technology remain the most likely route toward greater market adoption in larger installations."